Abstract | ||
---|---|---|
Precise prediction of users' next choices in time is critical for users' satisfaction and platforms' benefit. A user's next choice heavily depends on the user's long-term preference and recent actions. However, existing methods either ( 1) ignore the long-term personalized preference or the recent sequential actions of users, or ( 2) can't update the model in time when receiving users' new action information. To solve these problems, we propose an online personalized next-item recommendation method via long short term preference learning. The proposed method integrates the information of users' long-term personalized preference and short-term sequential actions to predict the next choices. The trained model could be updated online via an extra preference transition matrix. Experimental results on our real-world datasets show that the proposed method consistently outperforms several state-of-the-art methods. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1007/978-3-319-97304-3_70 | PRICAI 2018: TRENDS IN ARTIFICIAL INTELLIGENCE, PT I |
Keywords | Field | DocType |
Personalized recommendation, Sequential patterns, Online update, Collaborative filtering, Next-item prediction | Collaborative filtering,Computer science,Preference learning,Artificial intelligence,Next Choice,Machine learning | Conference |
Volume | ISSN | Citations |
11012 | 0302-9743 | 0 |
PageRank | References | Authors |
0.34 | 15 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yingpeng Du | 1 | 4 | 2.78 |
Hongzhi Liu | 2 | 88 | 14.92 |
Yuanhang Qu | 3 | 0 | 0.68 |
Zhonghai Wu | 4 | 34 | 12.36 |